• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

EVGA retracts PrecisionX 15 download - issues a statement

That you fail to understand the difference doesn't surprise me.



I think you'll find it's a forum full of people who also program, and know precisely what a **** move it is to stop paying someone but continue to use their software.

That you took that as a serious comment really, really doesn't surprise me at all :rolleyes:

As for a forum full of programmers, LOL, half the people here probably haven't written a piece of code since they were in school, the other half probably are still in school. (except for the fact that it is now the summer holidays)
 
Is it just me, can someone else clarify so that I don't get named a troll again, but is it namely the people that don't have use for Precision that are being overly negative towards EVGA?

Huge coincidence :)

Fascinating actually.

I have 2x570 Classys, 2x460s 2x480s and as of today 2 780TIs Classys all EVGA of course and will continue to buy EVGA but that doesn't stop me from calling a spade a spade like any honest rational person
 
So it's nothing to do with one guy potentially getting his life's work nicked by a huge company?

Well if they've nicked the code then MSI should call in their lawyers, but at the end of the day it's not like AB is locked to MSI products only. Can't see how this will affect me at all but in principal EVGA should compensate me with a 6Gb 780ti classified. I'm not asking for much :D
 
If no one can produce the offending lines of code and proof, I don't think anyone can accuse EVGA of any wrong doing and that includes Unwinder.

EVGA are not my favorite company but all I see here is people including Unwinder accusing them of stealing code without any proof.

If no one can provide the offending lines of code this really is a dead thread.

If someone can provide it then things may get interesting.
 
If no one can produce the offending lines of code and proof, I don't think anyone can accuse EVGA of any wrong doing and that includes Unwinder.

EVGA are not my favorite company but all I see here is people including Unwinder accusing them of stealing code without any proof.

If no one can provide the offending lines of code this really is a dead thread.

If someone can provide it then things may get interesting.

you can do it yourself

Unwinder said:
Boulard83 said:
This guy is try to sort out some of the lies.

http://forums.evga.com/FindPost/2192530


Originally Posted by ragingcain
Hey all, my binary analysis has made it's way over here.

I just want to state that disassembly/decompilation is not an exact science so I did go by other factors such as the size of the program and the potential for simple function call renames. I can not replicate Unwinder's claims. If someone else can please step up.
I don’t know how it is possible for programmer with at least a bit of reverse engineering experience to be unable to replicate the claims. I don't know how can a programmer misunderstand "string table" term and call it "Visual Studio 2008 IDE output of compilation". I don’t know why the binary analysis is applied to DLLs when I more than explicitly said that those things were traced inside EXE. I do hope it is just a question of skills.

Anyway, you can easily replicate the claims using the following simple steps:

- All string tables are located inside PrecisionX.exe. You can easily view this application in any binary file viewer (e.g. I use FAR) and view it as UNICODE (all string tables use UNICODE encoding). After opening it simply do a simple search for “RivaTuner” text (once again, in UNICODE). This way you’ll easily locate the string table, which belongs to original RivaTuner core with no doubts. For example in 5.0.0.16 (standalone version) it is located in close to 1E2xxx offset. It is also inside 5.0.0.17 as well. There are traces very specific to RT core, e.g. RivaTuner’s USF skin compiler/decompiler messages (“Failed to decompile %s skin”, “%s skin has been successfully decompiled” etc.), messages related to specific RivaTuner core components, e.g. RTCore driver loader messages, RivaTuner Task Scheduler Helper loader messages (“Failed to load RTTSH.dll”) and many more. Those things physically cannot be inside in-house project.
- You can verify original RT resources usage and easily find parts of original RTMUI localization engine in “Help” folder. Just open MAP file located in that folder as an ASCII text. This file is a part of original runtime translation engine, which allows RTMUI engine to map identifiers of controls to text files containing context help for corresponding controls. Those control identifiers are automatically generated by system when you create GUI, it is impossible to generate exactly the same IDs because original UI is normally being changed/modified during many years, some new controls are being added so IDs order is rather specific, etc, etc. So even if EVGA wanted to create 100% the same GUI and visually copy it, the order and IDs would be different. Those things uniquely identify original RivaTuner core GUI, even allow you to see how the controls were added, which of them were added first, etc, etc. Just compare contents of MAP file in new product with original one to see that it is exactly the same. Even missing G15 LCD output options are there:

Code:
; Properties \ Monitoring tab
ID_1015 = Properties\Monitoring\HARDWARE_POLLING_PERIOD
ID_1016 = Properties\Monitoring\HARDWARE_POLLING_PERIOD
ID_1076 = Properties\Monitoring\SOURCES_LIST
ID_1125 = Properties\Monitoring\SHOW_IN_OSD
ID_1077 = Properties\Monitoring\SHOW_IN_LCD
ID_1074 = Properties\Monitoring\LCD_FONT_COMBO
ID_1078 = Properties\Monitoring\SHOW_IN_TRAY
ID_1039 = Properties\Monitoring\TRAY_ICON_COLOR_PREVIEW
And once again, let’s just stop beating dead horse. It is time to move on, leave this stinky story behind and switch to more pleasant things now. Considering that new Precision uses almost 100% visual/functional clone of RT's hardware monitoring module and claimed to be designed in-house, I decided to start the real competition from AB side and let users see how original RT's hardware monitoring "done right" should look like. Since seeing PX 15 on Computex about couple months ago I was working on pumping up Afterburner's hardware monitoring module and powering it with more things from original RT. Here is the list of changes you can see in new version pretty soon. And let the true competition begin.


Version 4.0.0

• Various parts of hardware monitoring module have been pumped up to improve hardware monitoring usability and flexibility and leave competing tools back in the dust:
o Added layered monitoring graphs rendering mode. Now you may right click source graph in monitoring window, select “Attach” in the context menu then point to destination graph to attach source graph to it and create a group of layered graphs. This feature allows you to render as many layered graphs on the same grid as you wish. The colors of graphs in layered rendering mode can be customized independently of each other so you can easily identify them
o Added multi-column monitoring graphs rendering mode. Now you can adjust the number of graph columns in “Active monitoring graphs” section in “Monitoring” tab
o Added “Override graph name” option to “Monitoring” tab. Now you can rename the graphs displayed in hardware monitoring window
o Monitoring history buffer size is no longer defined by monitoring window width. Now pre-history buffer size is fixed and stores the last 3600 samples (1 hour for 1000ms polling period) for each graph
o Improved tray icon monitoring module:
o Now you can select either text mode or barchart indicator mode for each value displayed in tray icon. Barchart indicator mode can be extremely useful for visualizing data like GPU / CPU usage
o Improved Logitech keyboard LCD monitoring module:
o Ported to new Logitech API to provide support for newer Logitech LCD displays
o Added support for color LCD display of Logitech G19/G19s keyboards
o Added graph mode support for color LCD display of Logitech G19/G19s keyboards. Now in addition to previously available text mode you can optionally select graph mode and see exact copy of MSI Afterburner’s monitoring graphs displayed directly inside the keyboard LCD. You can also press “Menu” soft button on your Logitech G19/G19S keyboard to toggle between text and graph modes dynamically in realtime
o Added acceleration support to LCD scrolling implementation
o Added larger 8x12, 10x12, 12x12 and 12x16 fonts support for text mode
• Added “Regional settings” section to “User Interface” tab:
o Temperature format settings allow you to switch between Celsius and Fahrenheit format for monitored temperatures. Please take a note that this setting affects temperature readouts only. Hardware related temperature adjustments (e.g. fan speed to temperature mapping curve for all cards or temperature target adjustment for NVIDIA Kepler series) are always being displayed and adjusted in Celsius for maximum unification, safety and compatibility
o 12 hours / 24 hours time format settings allow you to configure time format for On-Screen Display and hardware monitoring window
• Added “Enable low-level IO driver” option to the “Compatibility properties” section in “General” tab
• Display device enumeration implementation has been modified slightly to allow monitoring Intel iGPUs when low-level IO driver is not enabled
• Improved handshaking algorithm reduces the risk of seeing multiple running instances of child processes (e.g. RTSS)
• Added automatic prerecording settings to “Videocapture” tab. When you enable automatic mode prerecording session is being started automatically on each 3D application startup. Please take a note that in this case you can still use video prerecord hotkey to stop then manually restart prerecording session if necessary
• Drastically improved skin engine:
o Improved skin compiler gives more detailed error messages when skin compilation fail due to error in some source image file
o Source image file format is no longer limited to 24-bit BMP files only. Now skin compiler supports all possible bit depths for BMP format and fully supports PNG format with alpha channel
o Added built-in bitmap effect for extracting alpha-channel from PNG image files
o Skin format has been upgraded to v1.3. New format supports alpha channel based transparency for skinned window, allowing skin designers to define semi-transparent skin areas, apply antialiasing to the skin window edges and so on
o Added new skinned window composition modes support and “Skin composition mode” settings to “User interface tab”. New settings allows you to use one of the following modes:
o Traditional mode – suits best for backward compatibility with existing skins and performance testing
o Layered mode with colorkey - provides much faster rendering of skins with non-rectangular window shape and additionally allows you to adjust transparency of skinned window
o Layered mode with alpha – provides per-pixel alpha channel support and advanced visual effects for compatible skins and also allows you to adjust transparency of skinned window
o Skin format reference guide has been updated to v1.7 to document these changes
o Full skins cross-compatibility with other overclocking applications based on RivaTuner engine. Special GUI transformation layer allows you to use the skins designed for third party RivaTuner based overclocking applications and makes the process of migration to MSI Afterburner from such overclocking tools much more comfortable for you. You can keep the look and feel of your preferred overclocking application and at the same time enjoy extended MSI Afterburner’s features including full range of supported graphics cards, industry leading powerful and robust monitoring module, flexible video recording features and many more
• RivaTuner Statistics Server has been upgraded to v6.1.3
 
Last edited:
If AMD copied and used GW's code, you would be baying for blood.;)

The Hulk would not stand a chance against that kind of rage :D:D:D:D:D. Me no like Greg Smash.

I think you have me completely misunderstood. I favor nVidia for sure but I don't wish to see my fellow gamers missing out, regardless of what brand is in their machine. So when you make baiting statements like that, you couldn't be further from the truth. AMD are more than welcome to copy GameWorks as far as I am concerned.
 
So what does all this mean, has anyone found any offending lines of code or not.

Or are they getting hung up with references and names.

It's all there in that quote..... There's a lot more than just few references that have been lifted.

- All string tables are located inside PrecisionX.exe. You can easily view this application in any binary file viewer (e.g. I use FAR) and view it as UNICODE (all string tables use UNICODE encoding). After opening it simply do a simple search for “RivaTuner” text (once again, in UNICODE). This way you’ll easily locate the string table, which belongs to original RivaTuner core with no doubts. For example in 5.0.0.16 (standalone version) it is located in close to 1E2xxx offset. It is also inside 5.0.0.17 as well.

- There are traces very specific to RT core, e.g. RivaTuner’s USF skin compiler/decompiler messages (“Failed to decompile %s skin”, “%s skin has been successfully decompiled” etc.), messages related to specific RivaTuner core components, e.g. RTCore driver loader messages, RivaTuner Task Scheduler Helper loader messages (“Failed to load RTTSH.dll”) and many more. Those things physically cannot be inside in-house project.

- You can verify original RT resources usage and easily find parts of original RTMUI localization engine in “Help” folder. Just open MAP file located in that folder as an ASCII text. This file is a part of original runtime translation engine, which allows RTMUI engine to map identifiers of controls to text files containing context help for corresponding controls. Those control identifiers are automatically generated by system when you create GUI, it is impossible to generate exactly the same IDs because original UI is normally being changed/modified during many years, some new controls are being added so IDs order is rather specific, etc, etc. So even if EVGA wanted to create 100% the same GUI and visually copy it, the order and IDs would be different. Those things uniquely identify original RivaTuner core GUI, even allow you to see how the controls were added, which of them were added first, etc, etc. Just compare contents of MAP file in new product with original one to see that it is exactly the same. Even missing G15 LCD output options are there
 
IIRC Unwinder said EVGA stopped paying royalties in November/December. If you look on Guru3D, it's pretty obvious from his posts he had a massive attitude towards them well before that. Not to mention the people that use it as well to boot.
 
IIRC Unwinder said EVGA stopped paying royalties in November/December. If you look on Guru3D, it's pretty obvious from his posts he had a massive attitude towards them well before that. Not to mention the people that use it as well to boot.

I cant see what difference any of that makes to this situation. You are assuming that his problem was with people using it and not people making demands to add features to precision which EVGA were not prepared to pay for.

Guess you must know something we dont. again.

It's amazing how people think they know something we all don't, isn't it James.
 
Last edited:
I think it's you that's doing a lot of the assuming...

I'm merely pointing out what he has posted in the past which clearly insinuates he wasn't happy long before any of this. Just because it's not something you can copy and past from the poorly handled Guru3D news story doesn't mean it's not relevant.

You're the one that seems to know everything.
 
It's not relevant, you and Kaap keep talking about entirely unrelated things.

EVGA can hate Unwinder and end their business relationship with him, no one gives a crap about that, unwinder doesn't appear to give a crap about that. That isn't the issue being talked about.

Regardless of when the relationship broke down, if it broke down, or why it broke down, none of that makes it okay to release a new version they say isn't using Unwinders work when it does. It doesn't matter if they did or didn't do this, I haven't taken a stance on this. But this is the ONLY issue "on the table" so to speak. unwinder says they are using his code with no basis to do so. EVGA say they aren't using his code, contracts are irrelevant(because EVGA have stated they have no right to use his code and are claiming to NOT be using his code), their relationship is irrelevant, how much he did or didn't hate users is irrelevant, how nice a person he is, is irrelevant, the temperature in Russia is equally irrelevant.

There is ONE discussion point, Unwinder says they are using his code, and EVGA say they aren't, both sides confirm EVGA have no legal basis to use his code.

Everyone making excuses for EVGA is bringing up anything but the actual issue at hand... not surprised.
 
Nobody is making excuses for EVGA.


Yet, dare I say what goes around? Or is that too unpolitical. Maybe we should get back to the facts we don't have yet. Given EVGAs vague responses, I think it's obvious to everyone that something is afoot that they'd probably rather not admit to. Before you jump down my throat, I've not once said EVGA are innocent, but as far as contractually we are taking his word for it.

http://forum.guru3d.com/showthread.php?p=4567060

More demonstration of his word
 
It's not relevant, you and Kaap keep talking about entirely unrelated things.

EVGA can hate Unwinder and end their business relationship with him, no one gives a crap about that, unwinder doesn't appear to give a crap about that. That isn't the issue being talked about.

Regardless of when the relationship broke down, if it broke down, or why it broke down, none of that makes it okay to release a new version they say isn't using Unwinders work when it does. It doesn't matter if they did or didn't do this, I haven't taken a stance on this. But this is the ONLY issue "on the table" so to speak. unwinder says they are using his code with no basis to do so. EVGA say they aren't using his code, contracts are irrelevant(because EVGA have stated they have no right to use his code and are claiming to NOT be using his code), their relationship is irrelevant, how much he did or didn't hate users is irrelevant, how nice a person he is, is irrelevant, the temperature in Russia is equally irrelevant.

There is ONE discussion point, Unwinder says they are using his code, and EVGA say they aren't, both sides confirm EVGA have no legal basis to use his code.

Everyone making excuses for EVGA is bringing up anything but the actual issue at hand... not surprised.

Why are you deciding what is relevant and what is not.

I still don't see where EVGA have used Unwinders work, all I see is how his and EVGA's parts work together.

Unwinder needs to say exactly where EVGA have nicked his work in the part of the code that both parties agree that EVGA have no claim to.

All I do see at the moment is Unwinder splitting hairs about the minor things.

I also think that you are not being objective and need to step back a bit rather than arguing tiny things with huge walls of text. james.miller also has a difference of opinion but to his credit he is making his point in a more objective way which is well worth reading.
 
Back
Top Bottom